Estimate of the Collins fragmentation function in a chiral invariant approach

We predict the features of the Collins function, which describes the fragmentation of a transversely polarized quark into an unpolarized hadron, by modeling the fragmentation process at a low energy scale. We use the chiral invariant approach of Manohar and Georgi, where constituent quarks and Goldstone bosons are considered as effective degrees of freedom in the non-perturbative regime of QCD. To test the approach we calculate the unpolarized fragmentation function and the transverse momentum distribution of a produced hadron, both of which are described reasonably well. In the case of semi-inclusive deep-inelastic scattering, our estimate of the Collins function in connection with the transversity distribution gives rise to a transverse single spin asymmetry of the order of 10%, supporting the idea of measuring the transversity distribution of the nucleon in this way. In the case of e+ e- annihilation into two hadrons, our model predicts a Collins azimuthal asymmetry of about 5%.